[bgolash]

This is an update on the machine I started in 2004. After using the machine I was never happy with the rigidly of the gantry. Sudden stops produced spindle shake. Two axes had Thk square linear motion elements. So no shake could be traced back to those axes. The X axis used Thompson 1 ¼ round linear motion elements. After contacting Thompson’s engineering team I had a potential recommendation. They mentioned I might have self-aligning bearings, which by their very nature have a looser tolerance to accommodate misalignments. After taking the machine apart it was confirmed that indeed the machine had the self-aligning versions.

It seems new replacement bearing would be nearly $100 apiece. Needless to say the added cost seemed too high when I had no idea what level of improvement could be seen. Ebay comes to the rescue. New old stock was available at a price of $29 per bearing. It was a bit of a process taking the machine apart to install the new bearing but in the end it was worth the time. The machine is now much improved.

I had contemplated selling the machine and starting a new machine. A few years ago I managed to get a Bridgeport milling machine. With the mill I would have been able to machine all the parts for a new machine. But with the improvements with the bearing upgrade I decided to just improve various parts of the machine. I had purchased a used nook xps 5/8 diameter for the original build. When I installed it I was aware it was not straight. So I decided to make new bearing blocks and purchase a new Nook ball screw and machine the ends. I have a South Bend lathe so machining began. I started by using carbide but that ended up being a problem. It just would not remove the material easily. This is a result of the casehardening of the screw. I remembered a Machinist recommending using high-speed steel tool bits. So none were ground so I had to grind my own. The improved metal removable was amazing with the newly ground tool bit. I got through the process and even single pointed the threads on the ball screw. I was very happy with the results.
.
Then I decided to upgrade the electronics. Originally I had my hands full dealing with the original control box. I managed to get through the process but thought it could be improved. Much of the cables were wired directly to the Geckos and BOB. I decided to do a new control box for the machine with panel mount cable connectors.
I managed to find an industrial control enclosure on eBay for $95.00. Since it was a second hand box the top and bottom were riddled with many holes. So I made 1/16 thick cover plates from aluminum and milled all the holes I needed on my Bridgeport. The Plexiglas front originally had a rectangular cut out. This required a new piece of Plexiglas installed for the repair. In addition, a side plate was made to allow the mounting of connectors. Years ago I ordered differential encoder cables from US digital. They came with non-panel mount ends. So I needed a way to make these connectors panel mount. The solution was to tap two 4-40 holes adjacent to the rectangle cutout and use 4-40 bolts and washers top and bottom to hold the connectors from being pushed in or out when installing or removing the matching connector.

Also removed all the hardware and sanded the box then I shot some Eastwood epoxy primer on the enclosure. Now it’s time for the wiring of the Gecko 320, limit switches etc. Part of the upgrade was removing some of the old limit switches from the machine. The machine now has all snap action limits instead of a mix of proximity and photo sensors. All new wire and new larger cable carriers were added to the machine. A Baldor 1 HP Dc motor and variable speed controller powers the spindle. Just to be safe I ran the spindle motor cable on the outside of the cable carrier to minimize its emi to the other cables within the cable carrier. The motor adds a good amount of weight to the Z axis but with the servos and a drive reduction I never see any servo faults. Something that was a concern during the original construction. One major benefit is the spindle that runs at maximum rpm of 5000 is very quiet when compared to a typical wood router.
New mounting brackets were needed to accommodate the new limit switches. Much of the control electronics were purchased from Automation Direct. They have excellent prices on many of the parts required. The power supply was purchased from Page not found | Automation Technology Inc Toroidal PSU (KL-7220). The old power supply was homemade with a 54 VDC capacity. My servos were purchased from Camtronics, inc. -- CNC with Dan Mauch. The new power supply is very close to the maximum allowable voltage of my motors. I assumed this increase in voltage would enhance the speed of the servos. Other parts are Bob Campbell’s breakout board, Ethernet Smooth Stepper and a c17 cnc4pc board, which monitors the servos for faults. In the original control box the grounding of the shields were an afterthought.

The new build concentrated on grounding. So far there are no strange noise issues. The box uses din rail power supplies and terminal blocks. It’s a very nice system and made wiring a lot easier. The control box is 24 x 30 inches. It ended up being a perfect size for all the component. I like the fact that it comes with a Plexiglass front. It’s nice to be able to check the various LEDs when the machine is running. It also forced me to do the wiring neater than the first control box. I`m also glad I found the box used on eBay. A new control box from this manufacturer is $500.
The next update will be a dust collection system for the machine. I assume it will need to be retractable to allow easy tool changes.

[CNCfun&games]

Hi Boglash,

Your original design was great!!
The new design which addressed the problems you encountered and the solutions implemented
is very interesting..Thanks for posting!
The electronic packaging is very clean & enviable. Keep up the good work.

W. Smith Mission Viejo, CA